Prostate cancer is the most common cancer in men and, after lung cancer, is the second leading cause of cancer deaths in men. Because a large number of prostate cancers remain localized and never spread, a critical challenge in the management of prostate cancer is trying to determine which patients should be treated aggressively; this challenge arises in part because of the lack of biomarkers that predict the biological behavior of prostate cancer. Ras and Rho are small G proteins involved in regulating cell growth, cell adhesion and motility, and apoptosis. In some cell systems, a constitutively- activated Ras is sufficient to induce cellular transformation and in animal tumor models, Rho clearly plays a role in cell migration and the development of metastases. We have developed methods to assess the activation states of Ras and Rho in human tissue and have found that Ras is highly activated, in the absence of a genetic ras mutation, in neuronal tumors and in breast and lung cancers. The basis for increased Ras activation appeared to be through overexpression of a growth factor receptor that signals through Ras, e.g., the epidermal growth factor (EGF) or ErbB-2 receptor, or through decreased activity of a RasGTPase. A mutated constitutively-activated EGF receptor is expressed in a high percentage of prostate cancers and the wild type EGF receptor and the ErbB-2 and ErbB-3 receptors appear to be overexpressed in some prostate cancers. In addition, prostatic stromal and epithelial cells produce a large number of growth factors and biopeptides leading to paracrine and autocrine stimulation of Ras and Rho. Thus, even though activating mutations in the ras gene are rare in prostate cancer, we hypothesize that Ras, as well as Rho, may be activated in prostate cancer and preliminary data on a small series of cancers support this hypothesis. The tumors with increased Ras and Rho activation appeared to exhibit a more malignant phenotype and we additionally hypothesize that high Ras or Rho activation may be a biomarker for a more aggressive tumor. We now propose to study two large series of prostate cancers, one prospectively and one retrospectively, in order to test these two hypotheses.